Open Access
Add to BibSonomyAbstract
We systematically study a type of plasmonic light absorber based on a monolayer of gold nano-spheres with less than 30 nm in diameters deposited on top of a continuous gold substrate. The influences of particle size, inter-particle distance, particle-substrate spacer size etc on the resonance are studied thoroughly with a 3D finite-element method. We identified that the high-absorption resonance is mainly due to gap plasmon (coupled through particle bodies) when the separation between neighboring nano-spheres is small enough, such as close to 1 nm; at larger particle separations, the resonance is dominated by particle dipoles (coupled through the host dielectric). Experimentally, an absorber was fabricated based on chemically-synthesized gold nanoparticles coated with silica shell. The absorber shows a characteristic absorption band around 810 nm with a maximum absorbance of approximately 90%, which agrees reasonably well with our numerical calculation. The fabrication technique can be easily adapted for devising efficient light absorbers of large areas.
© 2013 Optical Society of America
1. Introduction
Noble metals, like copper, silver, and gold are used for mirrors since ancient time. However, when they are patterned in subwavelength nanostructure, light reflection from such planar structure may disappear because light can be strongly coupled to the collective electron excitations and henceforth damped through collision with lattice and surfaces. This leads to efficient plasmonic light absorbers. The collective electron oscillation excited by optical-frequency electromagnetic wave is known as localized surface plasmon resonances (LSPR) [1]. LSPR is usually manifested as a characteristic absorption or extinction peak near to its resonant wavelength. This resonant frequency strongly depends on the corresponding nanostructure’s size, shape, and the surrounding environment [2]. Thanks to modern nanofabrication techniques such as electron-beam lithography (EBL) [3], focused-ion beam milling [4], or self-assembly of colloids [5], the applications of optical properties of LSPR have been explored enormously in recent years. For example, a lot of research has focused on utilizing the energy absorption process and photothermal effect [6] to enhance the efficiency of photovoltaic cells [7, 8] and thermoelectric devices [9]. Also, based on the resonance’s dependence on dielectric environment, Liu et al demonstrated an absorber which functions as a plasmonic sensor for measuring refractive index variations [10]. A near-infrared absorber based on a similar geometry was presented in [11]. The strong absorption of light achieved in [10, 11] is achieved by LSPR sustained by a layer of metallic nanoparticles (NPs) and a planar metal surface. However, in order for such a metal-insulator-metal (MIM) structure absorber [10, 11] to operate in the visible regime, its top-layer NPs are required to have a lateral size smaller than 100 nm [12]. Even with EBL process, it’s a formidable task to fabricate these tiny particles, especially when it comes to fabricate large samples. It was also shown that large-area light absorbers can be made with a thin continuous top-layer metal film instead of a layer of discrete metal particles [13]. However, such structures have relatively angle-dependent absorption bands. Recently, a large-area plasmonic absorber in the form of metal-dielectric nanocomposite showing almost 100% absorbance covering all the visible spectra range has been demonstrated by Hedayati et al [14]. It is fabricated by magnetron sputtering method which is cost-effective and scalable for large area devices.
Syntheses of metal NPs through wet chemistry approaches on the other hand have been well developed in recent years, mainly due to the advent of nanotechnology [15]. Such NPs, usually suspended in a liquid, can be prepared in large volumes. If they are deposited on a carefully prepared substrate, large-area absorber can be realized. Here, we study such a plasmonic absorber structure, which can be easily prepared with chemically synthesized gold nano-spheres. We numerically found that the absorber can be engineered to absorb light with a peak absorption at any wavelength from visible to 800 nm with a maximum absorbance larger than 70%. Additionally, the absorption is almost angle- and polarization-independent, which is desirable for solar energy harvesting devices. We also experimentally demonstrate one of these absorbers which has an absorption peak centered at 810 nm.
2. Numerical investigation
The absorber to be developed can be modeled by the idealized structure schematically shown in Fig. 1. A periodic array of gold nano-spheres, which are embedded in a silica layer, are distributed in a 2D triangular lattice on a continuous gold reflector. Very importantly, the nano-spheres are separated from the gold reflector by a spacer, which is a part of the silica host. The total thickness of the silica layer is t, the spacer thickness is s, and the gold reflector thickness is h. The interparticle separation is g, and the diameters of the spheres are d. In the following simulations we set t = 35 nm, s = 10 nm, h = 100 nm, d = 8 nm, and g = 1.8 nm unless otherwise specified. The 3D simulations are carried out by the finite element method using COMSOL Multiphysics. The optical constants of gold are from data measured by Johnson and Christy [16]; the refractive index of silicon dioxide is fixed at 1.5. The top-view of the unit cell used in our simulations is as shown later. The top and bottom sides of the unit are covered with perfectly matched layers; four sides are defined as periodic boundary conditions. We start with structures consisting identical spheres; the scenario with different sphere sizes will be briefly mentioned at the end of this section.
After extensive numerical case studies, we determined that the most critical geometrical parameter of this absorber is the particle-to-particle gap g. To show the effect of g on the resonance of the absorber, we calculate the absorption spectra for the d = 8 nm absorber at various g values for normal incidence case, as summarized in Fig. 2. Incoming light is polarized along x direction. All curves exhibit two distinctive absorption bands: one insensitive to g at the blue side of the visible spectrum, and one sensitive to g at the red side of the spectrum. The absorption at short wavelength is predominately from inherent material loss of gold due to interband electron transitions; the absorption band at longer wavelength is due to LSPR. It is shown that a decrease in g contributes to a red-shift of the LSPR resonant wavelength, and the amount of shift seems to increase with the decrease of g[17]. Since the nano-spheres are closely placed, the LSPR is certainly spectrally different from the dipole resonance originated from individual spheres. It would be interesting though to see how the LSPR resonance peak differs from the individual particle dipole resonance. For small spheres with diameter less than 10 nm, the phase retardation across the sphere can be neglected and quasi-static approximation can be applied. An individual sphere has its resonant wavelength satisfying the relation Re{εAu} +2εSiO2 = 0 [18], which gives rise to a resonant free-space wavelength of approximately 520 nm. While from Fig. 2 it is evident that by placing the spheres in close adjacency with each other, one can achieve a high-absorption band at a much longer wavelength.
Fig. 2 Calculated absorption spectra for d = 8 nm the absorber at various interparticle separation g at normal incidence.
To probe the nature of the resonance, we look into the field distribution of the absorber at the resonance. Figures 3(a) and 3(b) show the norm of electric displacement field for the g = 1 nm and g = 5 nm absorbers at their resonances, respectively. For the absorber with g = 1 nm [Fig. 3(a)], we see that the field is dominantly distributed in the gap between neighboring nano-spheres. Such a highly localized field confined in a dielectric gap between two metallic bodies usually referred to as a gap plasmon. Similar strong field enhancement in a dielectric gap between two metal tips was well discussed in works related to bowtie nanoantenna [19]. When the interparticle gap becomes larger [Fig. 3(b)], the field strength in the gap region weakens sharply, and the resonance is closer to resonance of individual particle dipoles. It is understandable that at other gap values, the resonance can be an equal mixture of the two above-mentioned resonances. It should be mentioned that as the structure has a periodicity of only ∼ 10 nm, each gap plasmon (particle dipole) in Fig. 3 is bonded with their neighboring gap plasmon (particle dipole). This is vital for understanding absorber structures with non-uniform nano-sphere inclusions.
Fig. 3 Calculated norm of electric displacement field (C/m2) at the cut plane parallel to the gold reflector and through the center of the nano-spheres for (a) g = 1 nm absorber at 680 nm wavelength and (b) g = 5 nm absorber with d = 8 nm at 560 nm wavelength for Ex polarized field at normal incidence in one unit cell.
Our investigation on the effect of the gap size stops at g = 1nm. Further increase of the LSPR resonance in wavelength by reducing g to a value less than 1 nm is possible but limited by quantum effects [20, 21], which are out of the scope of this work.
So far we have used a fixed sphere diameter d = 8 nm. In order to see the effect of changing d on the LSPR resonance wavelength, we simulated the absorption spectra of absorbers with a constant g = 1 nm whereas d is varying from 6 to 12 nm. The results are summarized in Fig. 4. It is seen that an increase in the diameter of the spheres shifts the resonant wavelength further to a larger value, slightly outside of visible spectrum but below 800 nm. One feature worth noticing is that when the diameter of nano-spheres gets bigger, the absorption peak becomes lower.
Fig. 4 Calculated absorption spectra for g = 1 nm absorber as a function of sphere diameter d at normal incidence.
Through the above analyses, we therefore conclude that by choosing g and d carefully, one can achieve high absorption at any wavelength from visible up to ∼ 800 nm.
In a recent experimental work a similar type of plasmonic absorber was fabricated [14]. The absorbers developed have a reflector at the bottom, a dielectric spacer in the middle, and a gold-nanoparticle layer on the top which is deposited through the magnetron sputtering process. The major difference is that there are “multiple layers” of gold particles randomly distributed in the dielectric host. In the reference it was argued that the broadband absorption of the structures is partially attributed to the magnetic-dipole resonance sustained by the top nano-composite and the bottom reflector, as was found in a MIM absorber structure [11]. However, through our simulations, we didn’t see any such magnetic-dipole resonance from 400 nm to even 1800 nm. Our analysis was also extended to absorbers consisting multiple layers of nano-spheres, and structures with several nano-spheres sizes.
In order to clearly identify the role of the reflector, we further compare the absorption spectra of the absorber for the structure with and without gold reflector (i.e. with air termination below SiO2). In the meantime, we change the spacer thickness s between the nano-sphere layer and the reflector. The diameters d are fixed as 8 nm, and the total thickness of the silica layer t = s+25 nm. As depicted by Figs. 5(a) and 5(b), the absorption spectra for structures with and without the gold reflector change periodically with an increasing s. In both Figs. 5(a) and 5(b), the additional spacer thickness necessary for achieving the maximum absorbance at the same wavelength position translates into an additional phase shift of 2π for the light in the spacer medium. Effectively the structure acts like a Fabry-Pérot Étalon with one reflector formed by the nano-sphere layer and the other by the bottom gold reflector or just the air termination. It is also seen that the maximum absorbance for the structure with gold reflector is about twice of that for the structure without gold reflector. Additionally, it is interesting to notice that the enhancement of absorption is particularly evident at the blue-side of the visible spectrum, where the absorption is mostly due to inherent material loss of gold.
Fig. 5 Calculated absorption spectra for the t = s + 25 nm, d = 8 nm, g = 1.8 nm absorber (a) with bottom gold reflector (b) without bottom gold reflector as a function of s at normal incidence.
The absorption characteristics of the nano-spheres based absorber is found to be insensitive to light incidence and almost completely independent of polarization. We performed a simulation with the angle of incidence changed from 0° to 80°. As shown in Fig. 6, the absorption profile doesn’t change much for incident angle up to 60°. This is desirable for, e.g. improving the efficiency of photovoltaic cells.
Fig. 6 Calculated absorption spectra for the d = 8 nm, g = 1.8 nm absorber as incident angle increases from 0° to 80°.
In the discussions above, we have only considered structure with identical spheres. However, it is in general not possible to fabricate nano-spheres of uniform size in practice. Refer to the unit cell shown in Fig. 3. Starting from a uniform-sphere structure with d = 10 nm and g = 1 nm, we reduce the center sphere’s diameter dm to 9 nm and then to 8 nm while keeping the corner spheres intact. All sphere positions are not changed. The absorption spectra of the three structures subject to a x-polarized normally-incident light are shown in Fig. 7. By changing the center sphere to a smaller size, we see that a new absorption peak arises on the left side of the main (“original”) LSPR absorption peak. The reduced center sphere diameter effectively introduces a new particle-particle gap size; the modified geometry, according to our previous discussion, effectively introduces a new absorption peak at smaller wavelength position. The “original” LSPR absorption peak is also shifted because the main LSPR and the newly-introduced resonance are not de-coupled. Finally in Fig. 7 we compare the spectra between the structure with dm = 8 nm and that with all sphere diameters at 8 nm (hence g = 3 nm). It is seen that the additional absorption peak in the nonuniform-sphere structure coincides with the single absorption peak of the uniform-sphere structure.
Fig. 7 Calculated absorption spectra for dm = 9 nm, g = 1 nm and dm = 8 nm, g = 1 nm absorber with nonuniform sphere size, in comparison with d = 10 nm, g = 1 nm and d = 8 nm, g = 3 nm absorber with identical sphere size.
3. Experimental realization
Based on the discussions in Section 2, we designed and fabricated an absorber as depicted in Fig. 8. The absorber was fabricated on a glass substrate, which was deposited with a 100 nm-thick gold and followed by a 3 nm-thick aluminium oxide using an electron beam physical vapor deposition. A 4 nm-thick Titanium layer was used between the gold layer and the glass substrate as an adhesive layer. Ideally in our design, a monolayer of silica-coated gold nano-spheres with gold sphere diameter at 20 nm and the silica shell thickness at 1 nm are placed on top in a triangular lattice. Such a structure exhibits LSPR-induced absorption bands similarly as the structure which we discussed in Section 2. The only difference is in the dielectric host. Practically such an absorber based on core-shell nano-spheres are much easier to realize.
Gold NPs were prepared chemically via a sol-gel method, by reducing 10 mM hydrochloroauric acid (HAuCl4) using 20 mM ascorbic acid and 1 mM sodium borohydride (NaBH4) in the presence of aqueous solution of cetyltrimethyl ammonium bromide (CTAB, 0.2 M) and silver nitrate (AgNO3, 2 mM) at room temperature. The mixture was stirred for one hour and kept at 25°C overnight. To coat the gold nanoparticles with silica shell, 1 mL of obtained suspension of gold NPs was diluted to 20 mL and the pH was tuned to ca. 12. When the temperature of this suspension was elevated to 70°C, 5 μL of tetraethyl orthosilicate (TEOS) was added, and the solution was collected after 1 h reaction and centrifuged to obtained silica-coated gold NPs (Au@SiO2).
The morphology of Au and Au@SiO2 NPs, as shown in Fig. 9, was characterized by JEM-2100F field emission transmission electron microscope (TEM) operating at accelerating voltage of 200 kV. The concentration of gold in colloidal suspensions was measured by inductively coupled plasma atomic emission spectroscopy (ICP-AES). The concentration of gold (element) and number of NPs in Au@SiO2 suspension are about, Au: 120 ppm (μg/mL); NPs: 1.505 × 1022/mL. In addition, the Au@SiO2 NPs in aqueous solution have a dark red appearance.
Then 170 μL Au@SiO2 are deposited on the coated substrate. After that, the sample is baked in a vacuum oven at 70°C for 2 hours. The SEM micrograph of the surface of the fabricated absorber is given in Fig. 10.
The setup for measuring reflectance is schematically shown in Fig. 11. A supercontinuum white light source (NKT SuperK Compact) with a repetition rate of 27 KHz is connected to a reflective collimator through a single-mode fiber. The collimated beam diameter is about 8.5 mm. The collimated beam then passes through a 600 μm pinhole. And then the beam is focused by an achromatic lens (f = 45 mm) on to the sample. On the back side of the sample, a 20× objective (f = 200 mm) and a CCD camera monitored with a PC are used to track the position of the beam on the sample. The reflected light was focused by another achromatic lens on the fiber core to couple the reflected light into a multi-mode fiber which is connected to the Optical Spectrum Analyzer (OSA).
The reflection (R) is measured using the system depicted above. The transmission can be neglected since the bottom gold reflector is much thicker than the skin depth in the concerned wavelength range. The absorbance (A) can be calculated by A = 1 − R. Figure 12(a) shows the measured absorption spectrum of the fabricated sample at 15° incident angle. A distinctive absorption peak around 810 nm wavelength is noticed. We also performed simulation with the idealized structure as given in Fig. 8. The calculated curve is superimposed in Fig. 12(a). The agreement between our simulation and experiment is reasonably good. We notice that the measured absorption value is in general higher than the simulated one, and the absorption band is also broader. This discrepancy should be due to the fact that the fabricated absorber is not perfectly flat at the scale of the focus beam size, which is limited by the current NP deposition process. The imperfect surface on one hand leads to random light scattering (therefore loss). On the other hand it would lead to a reflected beam with a slightly different (usually larger) shape, which in turn would cause the collected signal to be weaker. The size distribution of the NPs sampled based on 146 NPs is seen in Fig. 12(b). The average size is slightly more than 20 nm, which can be the reason why the measured absorption peak is at a longer wavelength than the calculated one (i.e. based on 20 nm nano-spheres). The broader absorption peak of the measured curve can be attributed to the inhomogeneity in particle size, shape and shell thickness, as expected.
Fig. 12 (a) Simulated and measured absorption spectra of the fabricated absorber at 15° angle of incidence; (b) particles sizes distribution based on 146 NPs as in Fig. 9(a).
4. Conclusion
In conclusion, we have numerically and experimentally investigated the absorption characteristics of absorbers that are easily producible at a large scale based on chemically synthesized gold-silica core-shell NPs. The highly efficient absorption band due to LSPR can be tailored to any wavelength from visible to 800 nm by carefully engineering the size of the gold particle and the separation between each particle. Our calculation shows that the high absorption band is not due to magnetic-dipole resonance between the NPs and the bottom gold layer. However, an inclusion of the bottom reflector, together with an appropriate spacer thickness, can indeed drastically increase the absorption efficiency of the overall structure. We also experimentally demonstrated an absorber based on Au@SiO2 core-shell NPs which has more than 50% average absorption from 600 nm to 900 nm. The maximum absorption value is measured at ∼ 90% centered at 810 nm. This approach effectively avoids expensive and tedious lithographic patterning process. Therefore it can be a practical and economical way for fabrication of, e.g. photo-thermo-electric energy conversion devices, provided that the light-induced heat can be efficiently converted to electricity.
Acknowledgments
This work is supported by the Swedish Research Council (VR) and VR’s Linnaeus center in Advanced Optics and Photonics (ADOPT). J. D. would like to acknowledge scholarship received under the European (Erasmus Mundus) Master of Science in Photonics programme.
References and links
1. M. L. Brongersma and P. G. Kik, Surface Plasmon Nanophotonics (Springer, P.O. Box 17, 3300 AA Dordrecht, The Netherlands, 2007) [CrossRef] .
2. K. L. Kelly, E. Coronado, L. L. Zhao, and G. C. Schatz, “The optical properties of metal nanoparticles: The influence of size, shape, and dielectric environment,” J. Phys. Chem. B 107, 668–677 (2003) [CrossRef] .
3. E. M. Hicks, S. Zou, G. C. Schatz, K. G. Spears, R. P. Van Duyne, L. Gunnarsson, T. Rindzevicius, B. Kasemo, and M. Käll, “Controlling plasmon line shapes through diffractive coupling in linear arrays of cylindrical nanoparticles fabricated by electron beam lithography,” Nano Lett. 5, 1065–1070 (2005) [CrossRef] [PubMed] .
4. M. Burresi, D. Diessel, D. v. Oosten, S. Linden, M. Wegener, and L. Kuipers, “Negative-index metamaterials: Looking into the unit cell,” Nano Lett. 10, 2480–2483 (2010) [CrossRef] [PubMed] .
5. C. J. Brinker, Y. Lu, A. Sellinger, and H. Fan, “Evaporation-induced self-assembly: Nanostructures made easy,” Adv. Mater. 11, 579–585 (1999) [CrossRef] .
6. X. Chen, Y. Chen, M. Yan, and M. Qiu, “Nanosecond photothermal effects in plasmonic nanostructures,” ACS Nano 6, 2550–2557 (2012) [CrossRef] [PubMed] .
7. H. A. Atwater and A. Polman, “Plasmonics for improved photovoltaic devices,” Nat. Mater. 9, 205–213 (2010) [CrossRef] [PubMed] .
8. K. Nakayama, K. Tanabe, and H. A. Atwater, “Plasmonic nanoparticle enhanced light absorption in gaas solar cells,” Appl. Phys. Lett. 93, 121904 –121904–3 (2008) [CrossRef] .
9. Y. Xiong, R. Long, D. Liu, X. Zhong, C. Wang, Z.-Y. Li, and Y. Xie, “Solar energy conversion with tunable plasmonic nanostructures for thermoelectric devices,” Nanoscale 4, 4416–4420 (2012) [CrossRef] [PubMed] .
10. N. Liu, M. Mesch, T. Weiss, M. Hentschel, and H. Giessen, “Infrared perfect absorber and its application as plasmonic sensor,” Nano Lett. 10, 2342–2348 (2010) [CrossRef] [PubMed] .
11. J. Hao, J. Wang, X. Liu, W. J. Padilla, L. Zhou, and M. Qiu, “High performance optical absorber based on a plasmonic metamaterial,” Appl. Phys. Lett. 96, 251104 (2012) [CrossRef] .
12. A. Tittl, P. Mai, R. Taubert, D. Dregely, N. Liu, and H. Giessen, “Palladium-based plasmonic perfect absorber in the visible wavelength range and its application to hydrogen sensing,” Nano Lett. 11, 4366–4369 (2011) [CrossRef] [PubMed] .
13. M. Yan, “Metal-insulator-metal light absorber: a continuous structure,” J. Opt. 15, 025006 (2013) [CrossRef] .
14. M. K. Hedayati, M. Javaherirahim, B. Mozooni, R. Abdelaziz, A. Tavassolizadeh, V. S. K. Chakravadhanula, V. Zaporojtchenko, T. Strunkus, F. Faupel, and M. Elbahri, “Design of a perfect black absorber at visible frequencies using plasmonic metamaterials,” Adv. Mater. 23, 5410–5414 (2011) [CrossRef] [PubMed] .
15. L. M. Liz-Marzán, M. Giersig, and P. Mulvaney, “Synthesis of nanosized gold-silica core-shell particles,” Langmuir 12, 4329–4335 (1996) [CrossRef] .
16. P. B. Johnson and R. W. Christy, “Optical constants of the noble metals,” Phys. Rev. B 6, 4370–4379 (1972) [CrossRef] .
17. H. Xu, E. J. Bjerneld, M. Käll, and L. Börjesson, “Spectroscopy of single hemoglobin molecules by surface enhanced raman scattering,” Phys. Rev. Lett. 83, 4357–4360 (1999) [CrossRef] .
18. S. Maier, Plasmonics: Fundamentals And Applications (Springer, 2007).
19. S. Kim, J. Jin, Y.-J. Kim, I.-Y. Park, Y. Kim, and S.-W. Kim, “High-harmonic generation by resonant plasmon field enhancement,” Nature 453, 757–760 (2008) [CrossRef] [PubMed] .
20. C. Cirac’i, R. T. Hill, J. J. Mock, Y. Urzhumov, A. I. Fernández-Domínguez, S. A. Maier, J. B. Pendry, A. Chilkoti, and D. R. Smith, “Probing the ultimate limits of plasmonic enhancement,” Science 337, 1072–1074 (2012) [CrossRef] .
21. K. J. Savage, M. M. Hawkeye, R. Esteban, A. G. Borisov, J. Aizpurua, and J. J. Baumberg, “Revealing the quantum regime in tunnelling plasmonics,” Nature 491, 574–577 (2012) [CrossRef] [PubMed] .
